Adaptor Proteins and Ras Synergistically Regulate IL-1-Induced ADAMTS-4 Expression in Human Chondrocytes

This information is current as Rasheed Ahmad, Judith Sylvester, Mushtaq Ahmad and of September 24, 2021. Muhammad Zafarullah J Immunol 2009; 182:5081-5087; ; doi: 10.4049/jimmunol.0803544 http://www.jimmunol.org/content/182/8/5081 Downloaded from

References This article cites 63 articles, 15 of which you can access for free at: http://www.jimmunol.org/content/182/8/5081.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 24, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts Errata An erratum has been published regarding this article. Please see next page or: /content/185/1/769.full.pdf

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2009 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Adaptor Proteins and Ras Synergistically Regulate IL-1-Induced ADAMTS-4 Expression in Human Chondrocytes1

Rasheed Ahmad,* Judith Sylvester,* Mushtaq Ahmad,† and Muhammad Zafarullah2*

Aggrecanases (a dystrophin and metalloproteinase with thrombospondin motif, ADAMTSs) are principal proteases involved in cartilage extracellular matrix aggrecan degradation. The role and relative contribution of MyD88, IRAK1, and TRAF6 adaptor proteins in IL-1␤ regulation of aggrecanase-1 (ADAMTS-4) is unknown. By small interfering RNAs-mediated knockdown, we show that IL-1␤-induced up-regulation of ADAMTS-4 in chondrocytes requires MyD88, IRAK1, and TRAF6 adaptor proteins. However, partial inhibition of ADAMTS-4 induction by their knockdown suggested the involvement of additional signaling proteins. Because IL-1␤ is also known to induce reactive oxygen species (ROS) through Ras-mediated activation of NADPH oxidase, we investigated the implication of Ras in ADAMTS-4 regulation. Ras knockdown, or inhibition of ROS by antioxidants ␤ along with the ablation of MyD88, IRAK1, or TRAF6 more potently down-regulated IL-1 -induced ADAMTS-4. In addition, Downloaded from IL-1␤-induced phosphorylation of downstream effectors, I␬B kinase ␣␤,I␬B␣, and activation of transcription factor NF-␬B was significantly reduced in the MyD88-, IRAK1-, TRAF6-, or Ras-deficient cells. The combined knockdown of Ras and individual adaptor proteins strongly blocked the activation of IKK␣␤,I␬B␣, and NF-␬B. These findings suggest that Ras, ROS along with MyD88, IRAK1, or TRAF6 synergistically mediate ADAMTS-4 regulation by IL1-␤. Thus, complete ablation of ADAMTS-4 induction could be achieved by combined inhibition of Ras and individual adaptor proteins, which may be of therapeutic value

in arthritis. The Journal of Immunology, 2009, 182: 5081–5087. http://www.jimmunol.org/

ntegrity of cartilage is essential for weight bearing ability, gies including neutralizing Abs and receptor antagonists are mobility, and flexibility of joints. Chondrocytes synthesize aimed at inhibiting TNF-␣ and IL-1 actions (7, 8). IL-1 binds to I and maintain cartilage extracellular matrix (ECM),3 which its receptor, TLR/IL-1R1, leading to complexation with IL-R is principally made up of type II collagen and proteoglycan accessory protein, recruitment of MyD88 intracellular adaptor aggrecates or aggrecan whose function is to resist compression. protein, interaction with IL-1R-associated kinase (IRAK), its Aggrecan monomers consist of a core protein from where chon- phosphorylation, and association with TNF receptor-associated droitin sulfate and keratan sulfate chains protrude. Interaction factor 6 (TRAF6). These events lead to activation of multiple of aggrecan hyluronon with link protein constitutes large ag- cascades including ERK, p38, JNK, AP-1, and NF-␬B pathways by guest on September 24, 2021 gregates (1). Due to their ability to inhibit ECM synthesis (2), (7, 9). ␣ proinflammatory cytokines such as IL-1 and TNF- produced Upon IL-1 exposure, aggrecan is destroyed first, followed by by inflammatory cells are considered as prime instigators of the digestion of collagen fibrils. Matrix metalloproteinase-13 cartilage ECM degradation during the pathogenesis of rheuma- (MMP-13 or collagenase-3) preferentially cleaves type-II col- toid arthritis (RA) and osteoarthritis (OA) (3, 4). Levels of IL-1 lagen while a family of aggrecanases or ADAMTSs (a dystro- are augmented in the synovial fluid and cartilage of patients phin and metalloproteinase with thrombospondin motif) is in- with arthritis (5, 6) and some of the major anti-arthritic strate- volved in physiological remodeling and pathological cleavage of aggrecan at the Glu373-Ala374 bond of interglobular domain *Department of Medicine, University of Montreal and Centre de Recherche du Centre of the core protein (10, 11). ADAMTS-4 is expressed in active Hospitalier de l’Universite´ de Montreál, Notre-Dame Hospital, Montreal, Quebec, form in OA cartilage and aggrecan fragments are found in OA Canada; and †Cardiovascular Research Institute (CVRI), Morehouse School of Med- icine, Atlanta, GA 30310 synovial fluid (12, 13). ADAMTS-5 and not ADAMTS-4 Received for publication October 22, 2008. Accepted for publication February knockout resulted in inhibition of cartilage aggrecan destruc- 9, 2009. tion, suggesting ADAMTS-5 as the predominant aggrecanase in The costs of publication of this article were defrayed in part by the payment of page mice (14, 15). However, ex vivo RNA interference experiments charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. with the human cartilage explants revealed that both ADAMTS-4 and Ϫ5 are required for human cartilage aggrecan degradation (16). 1 This work was supported by a grant (to M.Z.) and a fellowship (to R.A.) from the Canadian Institutes of Health Research and start-up funds (to M.A.) from Morehouse Blocking aggrecanolysis in the interglobular domain by genetic School of Medicine. means resulted in abrogation of cartilage erosion (17). Although 2 Address correspondence and reprint requests to Dr. M. Zafarullah, K-5255 Mail- implication of protein kinase C in aggrecanase has been dem- loux, Hoˆpital Notre-Dame du Centre Hospitalier de l’Universite´ de Montreál, 1560 Sherbrooke est, Montreál, Que´bec, Canada H2L 4M1. E-mail address: onstrated (18, 19), the detailed mechanism of IL-1 signal trans- [email protected] duction leading to ADAMTS-4 gene induction and the role of 3 Abbreviations used in this paper: ECM, extracellular matrix; RA, rheumatoid adaptor proteins in particular is not known. By RNA inter- arthritis; OA, osteoarthritis; ADAMTS, A dystrophin and metalloproteinase with ference-mediated genetic knockdown, we demonstrate that thrombospondin motif; IRAK1, IL-1R-associated kinase; TRAF6, TNF receptor- associated factor 6; siRNA, small interfering RNA; ROS, reactive oxygen species; MyD88, IRAK-1, and TRAF6, as well as Ras-driven reactive IKK, I␬B kinase; MMP-13, matrix metalloproteinase-13; NDGA, nordihydroguai- oxygen species (ROS) culminating on NF-␬B transcription fac- aretic acid. tor synergistically regulate IL-1-induced ADAMTS-4 gene Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00 expression. www.jimmunol.org/cgi/doi/10.4049/jimmunol.0803544 5082 AGGRECANASE-1 REGULATION BY MYD88, IRAK1, TRAF6 AND RAS Downloaded from http://www.jimmunol.org/

FIGURE 1. siRNA-mediated knockdown of MyD88, IRAK1, and TRAF6 expression results in the inhibition of IL-1␤-induced ADAMTS-4 gene expression and enzyme activity. A, Chondrocytes were transfected with either adaptor-specific siRNA or control siRNA. Cells were harvested 48 h later and protein lysates analyzed for MyD88, IRAK1, TRAF6, TRAF2, and ␤-actin expression by immunoblotting. Representative Western blots show specific inhibition of the target proteins. B, Control siRNA-transfected as well as MyD88-, IRAK1-, TRAF6-, and TRAF2-deficient chondrocytes were incubated with either IL-1␤ or TNF-␣. Supernatants and chondrocytes were harvested 24 h posttreatment, respectively, for total RNA extraction and ADAMTS-4 by guest on September 24, 2021 activity. ADAMTS-4 gene expression was examined by RT-PCR. C, ADAMTS-4 enzyme activity from the supernatants was determined by ELISA. The values are mean Ϯ SD of three separate experiments.

Materials and Methods Western blotting Chondrocyte cultures and antioxidant treatments Chondrocytes were harvested and incubated for 30 min with lysis buffer The normal primary human knee articular chondrocytes (Cambrex) were (Tris 62.5 mM (pH 7.5), 1% Triton X-100, 10% glycerol). The lysates were grown to confluence as high-density passage 2 monolayer cultures in Dif- then centrifuged at 14000 rpm for 10 min and the supernatants were collected. Protein concentration in the lysates was measured by the Bio-Rad ferentiation Bullekit medium where they maintain their differentiated phe- ␮ notype. These cells express cartilage-specific type II collagen marker as Protein Assay kit. Protein (20 g) samples were mixed with sample load- analyzed by Western blot analysis. The chondrocytes were grown in six- ing buffer, heated for 5 min at 95°C and separated by SDS-PAGE. Cellular well plates in DMEM (Invitrogen) with 10% FCS. Chondrocytes were proteins were transferred to Immobilin-P membrane by electroblotting. The washed with PBS, kept in serum-free DMEM for 24 h, and preincubated membranes were then blocked with 5% nonfat milk in PBS for 1 h, incu- with Trolox (100 ␮M) or nordihydroguaiaretic acid (NDGA) (20 ␮M) bated with the desired primary Ab diluted either in 5% milk in PBS or 5% BSA overnight at 4°C. For monitoring the efficiency of knockdown, protein (Calbiochem) for 1 h and then treated with either 0.1% BSA (control) or ␮ stimulated with IL-1␤ or TNF-␣ (10 ng/ml each) (R&D Systems) for 24 h. extracts were probed with MyD88, IRAK1 (2 g/ml, from ProSci), TRAF6 ADAMTS-4 mRNA levels were measured by RT-PCR and several pro- (1/200 dilution, Santa Cruz Biotechnology), TRAF2 (1/500 dilution, Cell teins by Western blotting as described below. Signaling Technology) and Ras (Upstate Biotechnology) Abs. The same or duplicate blots were subsequently used for measuring the levels of ␤-actin Small interfering RNA (siRNA) transfections (Sigma-Aldrich) loading control. Abs against phosphorylated I␬B kinase (IKK) ␣␤,I␬B␣, and total I␬B␣ were from Cell Signaling Technology and Chondrocytes were transfected with different siRNAs with a modified were used at 1/1000 dilutions. The blots were then washed four times with high-efficiency calcium phosphate precipitation method (20). In brief, tryp- TBS and incubated for 2 h with HRP-conjugated secondary Ab. Immuno- sin-released chondrocyte cell suspensions were incubated with the respec- reactive bands were developed using an ECL substrate (Amersham Bio- tive siRNAs or equivalent amount of negative control (200 nM)-calcium sciences) and visualized by autoradiography. phosphate precipitate for 30 min and plated in the serum-containing medium for 3 h. siRNAs were from the following sources: MyD88 Validated RT-PCR Stealth Duopak siRNA and control siRNA (Invitrogen), Silencer Negative Control no.2 siRNA, IRAK1 Silencer Validated siRNA (Ambion), Control Total RNA was analyzed for measuring the ADAMTS-4 and GAPDH siRNA (sc-44236), TRAF6 siRNA (sc-36717), and TRAF2 (sc-29509) mRNA levels by RT-PCR with ADAMTS-4-specific primers (21) yielding from Santa Cruz Biotechnology were also used. Ras siRNA and control 692 and 226 bp cDNA bands. The amplification profile was one cycle at siRNA (Ras siRNA Assay Kit) were from Upstate Biotechnology. Cells 94°C for 1 min, 35 cycles of 94°C for 1 min, hybridization at 60°C for 2 were washed with PBS, allowed to recover in serum-containing medium, min, and extension at 72oC for 3 min, followed by one extension cycle of maintained in serum-free medium, and then stimulated with IL-1␤ or 7 min at 72°C. The PCR was performed in a DNA cycler (R&D Systems) TNF-␣. An equal amount (20 ␮g) of protein was analyzed for measuring ina50␮l reaction with 1.25 mM dNTPs, TaqDNA polymerase, and re- the respective protein levels. spective primers. Aliquots of 10 ␮l from the 50 ␮l PCR were analyzed on The Journal of Immunology 5083 Downloaded from http://www.jimmunol.org/

FIGURE 2. Ras is required for IL-1␤-induced ADAMTS-4 gene expression. A, Chondrocytes were transfected with Ras siRNA or control siRNA (200 nM). Cell lysates were harvested at 48 h after transfection and inhibition of Ras expression was detected by immunoblotting. B, Ras-deﬁcient chondrocytes were treated with IL-1␤ or TNF-␣ for 24 h and ADAMTS-4 mRNA analyzed by RT-PCR and activity by ELISA. C, Chondrocytes ␤

were transfected with Ras siRNA alone or in combination with MyD88, IRAK1, or TRAF6 siRNAs and were then stimulated with IL-1 for 24 h. by guest on September 24, 2021 Total chondrocyte RNA was analyzed for ADAMTS-4 mRNA by RT-PCR and supernatants for ADAMTS-4 enzyme activity by ELISA. Data are expressed as mean Ϯ SD from three independent experiments. D, Control or adaptor proteins-deﬁcient chondrocytes were pretreated with Trolox (100 ␮M) or NDGA (20 ␮M) and then exposed to IL-1␤ for 24 h. RT-PCR and ELISA were used to determine ADAMTS-4 gene expression and enzyme activity, respectively.

1.4% agarose gels to detect ADAMTS-4 and GAPDH cDNA amplification All the experiments were performed at least three times and the reported products. results were reproducible. ELISA for ADAMTS-4 enzyme activity Results For measuring the biological activity of the ADAMTS-4 enzyme, super- MyD88, IRAK1, and TRAF6 are partially involved in natants were collected from the cells cultured in the six-well plates and IL-1␤-induced ADAMTS-4 up-regulation in chondrocytes Ϫ were kept at 20°C until analysis. For quantitative determination of ag- ␤ grecanase activity, Sensitive Aggrecanase Activity ELISA (Cat no. SEN- To determine whether IL-1 -induced-ADAMTS-4 expression is AGG-96, MD Biosciences) was used according to their protocol. The ag- mediated by IL-1R-associated adaptor proteins, siRNA technology grecanase in supernatant cleaves the recombinant interglobulin domain of was used to knock down the adaptor proteins, MyD88, IRAK-1, aggrecan, releasing a peptide with the N-terminal sequence of ARGSVIL and TRAF6. Chondrocytes were transiently transfected with either that is quantified by two anti-peptide mAbs (22, 23). The absorbance val- an unrelated siRNA (control siRNA) or gene-specific siRNA. Im- ues were read at 450 nm. munoblot analysis revealed the effective suppression of constitu- NF-␬B DNA binding assay tive MyD88, IRAK1, and TRAF6 protein expression in chondro- Chondrocytes were transfected with the siRNAs as above and then stim- cytes transfected with the specific gene-targeted siRNA (Fig. 1A). ulated with IL-1␤ or TNF-␣ for 60 min. For nuclear protein isolation, In contrast, nonspecific control siRNA did not inhibit the expres- NucBuster Protein Extraction Kit (Novagen) was used according to the sion of these adaptor proteins (Fig. 1A). These results demonstrate manufacturer’s instructions. Nuclear pellets were resuspended in 50 ␮lof the specificity and effectiveness of the knockdown approach. complete lysis buffer provided by the Active Motif kit. After 30 min in- IL-1␤ treatment of cells resulted in remarkable up-regulation of cubation on ice, samples were centrifuged and proteins measured with the Bio-Rad Protein Assay kit. NF-␬B DNA binding activity was determined ADAMTS-4 mRNA levels and biological activity (Fig. 1, B and by using TransAM NF-␬B ELISA Kit (Active Motif) as recommended by C). These ADAMTS-4-inducing effects of IL-1␤ were partially the manufacturer’s instructions. In brief, 4 ␮g of nuclear proteins were blocked in chondrocytes transfected with MyD88-specific siRNA added to each well of a 96-well plate precoated with immobilized oligo- (Fig. 1, B and C, left panels). These results demonstrate that IL- nucleotide containing the NF-␬B consensus site (5Ј-GGGACTTTCC-3Ј) ␤ followed by the addition of primary transcription factor Ab and secondary 1 -induced activity of the ADAMTS-4 gene requires expression HRP-conjugated Ab. Later, HRP substrate was added and colorimetric val- of MyD88 adaptor protein. In contrast, silencing of TRAF2, the ues measured at 450 nm were plotted as bar graphs. adaptor protein for TNF-␣ (24) did not affect IL-1␤-mediated 5084 AGGRECANASE-1 REGULATION BY MYD88, IRAK1, TRAF6 AND RAS

up-regulation of ADAMTS-4 expression (Fig. 1B). We also confirmed that siRNA for MyD88 did not reduce TNF-␣-induction of ADAMTS-4 (Fig. 1, B and C). To further decipher the signaling components downstream of MyD88, chondrocytes with knockdown of IRAK1 and TRAF6 were developed by using adaptor- targeted siRNAs. As shown in Fig. 1, B and C, cells deficient in either IRAK1 or TRAF6 displayed a remarkably diminished IL-1␤ induction of ADAMTS-4, which indicates the partial requirement of these adaptor molecules for the IL-1 signaling pathway leading to ADAMTS-4 gene expression in chondrocytes. As a negative control, we also confirmed that IRAK1- or TRAF6-deficient chondrocytes did not block the TNF-␣-induced activation of ADAMTS-4 gene expression and enzyme activity (Fig. 1, B and C). Combined knockdown of Ras along with MyD88, IRAK, or TRAF6 potently down-regulates IL-1␤ induced ADAMTS-4 expression From the experiments of previous section, we learned that

knockdown of individual adaptor proteins partially suppressed Downloaded from IL-1␤-induced ADAMTS-4 up-regulation. This led to the hypoth- esis that IL-1␤ could activate other signaling pathways to induce ADAMTS-4 expression. Because IL-1␤ can induce conversion of inactive GDP Ras into active GTP Ras (25, 26) as well as ROS (superoxide and hydrogen peroxide) through Ras-mediated activa- FIGURE 3. Impaired IL-1␤-induced activation of IKK␣␤,I␬B␣, and

tion of NADPH oxidase (27–29), Ras may therefore play a key NF-␬B in MyD88-, IRAK1-, or TRAF6-deficient chondrocytes. A, Cells http://www.jimmunol.org/ role in the regulation of IL-1␤-induced ADAMTS-4. To investi- were first made adaptor proteins-deficient by specific siRNA transfection gate the involvement of Ras in the up-regulation of ADAMTS-4, and subsequently stimulated with IL-1␤, or TNF-␣ for 20 min. Control we silenced Ras expression by using Ras-specific siRNA as con- chondrocytes were also treated with 0.1% BSA (control), IL-1␤, or TNF-␣ firmed by Ras immunoblot analysis. Ras protein levels were con- alone. The levels of phosphorylated IKK␣␤ and I␬B␣ were determined by siderably reduced in Ras siRNA-transfected cells compared with Western blotting of whole-cell lysates using specific Abs. The correspond- ␤ control siRNA-transfected chondrocytes (Fig. 2A). Ras deficiency ing Western blot for the actin levels is depicted in the bottom panel. B, Control siRNA-transfected or adaptor proteins-deficient chondrocytes were partially inhibited the induction of ADAMTS-4 mRNA and en- ␤ ␣ ␤ ␣ treated with IL-1 or TNF- for 1 h. Nuclear extracts were analyzed for zyme activity by either IL-1 or TNF- (Fig. 2B). In view of the NF-␬B DNA binding activity by ELISA and values measured at 450 nm by guest on September 24, 2021 observed partial inhibition of ADAMTS-4 expression and activity were plotted. The values are mean Ϯ SD of three separate experiments. by either Ras or by different adaptor proteins, we studied the effect of their combined ablation by transfecting chondrocytes with Ras siRNA and siRNA for MyD88, IRAK, or TRAF6. As depicted in tively blocked IL-1␤-induced phosphorylation of IKK␣␤ and Fig. 2C, we observed almost total inhibition of ADAMTS-4 in- I␬B␣ (Fig. 3, A and B). In contrast, siRNA for TNF-␣-related duction in chondrocytes transfected with combined Ras and adap- adaptor, TRAF-2, did not block IL-1-stimulated phosphorylation tor protein siRNAs. of IKK␣␤ and I␬B␣. However, TRAF-2 siRNA completely elim- Because superoxide can activate Ras (30), and Ras-driven acti- inated TNF-␣-induced phosphorylation of these mediators (Fig. 3, vation of NADPH oxidase leads to ROS (superoxide and hydrogen A and B, last lane). In contrast, MyD88, IRAK1, or TRAF6 block- peroxides) production (31), and 5-lipoxygenase (which catalyzes ade showed minimal effect on TNF-␣-activated phosphorylation of insertion of molecular oxygen in arachidonic acid during leukotri- IKK␣␤ and I␬B␣ (Fig. 3A). ene biosynthesis) is a source of ROS production (32), we examined IL-1␤ is known to stimulate phosphorylation of IKK␣␤ and the effects of well-documented antioxidants, Trolox (a vitamin I␬B␣ that leads to activation of the transcription factor NF-␬B E analog), and NDGA (a 5-lipoxygenase inhibitor) (32–35) on (36, 37). Therefore, we examined the IL-1␤-stimulated NF-␬B ADAMTS-4 induction. Pretreatment of chondrocytes with DNA-binding activity in the MyD88-, IRAK1, or TRAF6-defi- Trolox or NDGA followed by IL-1␤ stimulation partially cient chondrocytes. Consistent with the observed effect on in- blocked cytokine-induced activation of ADAMTS-4 mRNA and termediary proteins phosphorylation, IL-1␤-induced NF-␬B enzyme activity (Fig. 2D). Trolox or NDGA by itself had no DNA binding activity was significantly reduced in MyD88-, effect on ADAMTS-4 gene expression. However, IL-1␤ induc- IRAK1-, and TRAF6-deficient chondrocytes compared with the tion of ADAMTS-4 mRNA and activity was almost completely control siRNA-transfected cells (Fig. 3B). The MyD88, IRAK1, blocked in MyD88-, IRAK1-, or TRAF6-deficient chondrocytes or TRAF6 siRNAs had negligible effects on NF-␬B activity in cotreated with Trolox or NDGA (Fig. 2D). response to TNF-␣. However, TRAF-2 silencing did affect TNF-␣-mediated NF-␬B activity (Fig. 3B). Deficiency of adaptor proteins impairs IL-1␤-induced activation of IKK␣␤,I␬B␣, and NF-␬B Combined knockdown of Ras and adaptor proteins additively ␤ ␣␤ ␬ ␣ IKK␣␤ and I␬B␣ are important mediators of IL-1R signal trans- down-regulates IL-1 -induced activation of IKK ,I B , and ␬ duction leading to NF-␬B activation. To investigate the impact of NF- B adaptor proteins on these downstream IL-1␤ signal transducers, As IKK␣ and I␬B␣ are important intermediary molecules in IL1-␤ MyD88-, IRAK1-, or TRAF6-deficient chondrocytes were treated signal transmission, we investigated whether Ras deficiency af- with IL-1␤ for 20 min and cell lysates analyzed by Western fected phosphorylation of these proteins. To test this, Ras-deficient blotting. MyD88-, IRAK1-, and TRAF6-siRNA transfection effec- chondrocytes were incubated either with IL-1␤ or with TNF-␣ for The Journal of Immunology 5085

FIGURE 4. Ras siRNA in combination with MyD88, IRAK, or TRAF6 siRNAs diminishes IL-1␤-induced activation of IKK␣␤,I␬B␣, and NF-␬B. A, Chondrocytes were transfected with Ras siRNA and subsequently stimulated with IL-1␤ or TNF-␣ for 20 min. The cell lysates were analyzed by immunoblotting with the respective Abs. B, Control and Ras-deﬁcient chondrocytes were treated with IL-1␤ or TNF-␣ for 1 h. Nuclear extracts were subjected to ELISA for determination of NF-␬B DNA binding activity. C, Chondrocytes were transfected with Ras siRNA in combination with MyD88, IRAK1, or TRAF6 siRNAs and subsequently exposed to IL-1␤ or TNF-␣ for 20 min. Cell lysates were analyzed by immunoblotting with the indicated Abs. Downloaded from D, Chondrocytes deﬁcient in Ras along with MyD88, IRAK1, or TRAF6 were treated with IL-1␤ or TNF-␣ for 1 h and nuclear extracts were subjected to ELISA for measuring the NF-␬B DNA binding activity. The values are mean Ϯ SD of

three separate experiments. http://www.jimmunol.org/

20 min and cell lysates used for Western blot analysis. As expected ence technology. Partial inhibition of ADAMTS-4 expression pro- both IL-1␤ and TNF-␣ induced I␬B␣ phosphorylation with a con- vided strong and unprecedented evidence for the involvement of comitant increase in I␬B␣ degradation at 20 min. Ras knockdown MyD88, IRAK1, and TRAF6 in ADAMTS-4 regulation by IL-1␤. by Ras-specific siRNA (Fig. 4A) decreased IKK␣␤ and I␬B␣ This mechanism appears to be specific for IL-1␤, as knockdown of phosphorylation and increased I␬B␣ stability compared with the these proteins did not affect induction of ADAMTS-4 by TNF-␣. cells transfected with control siRNA (Fig. 4A). Because IL-1␤- The latter cytokine induces ADAMTS-4 primarily through TRAF2 by guest on September 24, 2021 induced phosphorylation of IKK␣␤ and I␬B␣ leads to subsequent adaptor protein, which transduces TNF-␣ signal (24). Similar pat- activation of NF-␬B, we also determined whether NF-␬B activa- terns of ADAMTS-4 mRNA and enzyme activities indicate that tion was affected by Ras deficiency. Chondrocytes deficient in Ras ADAMTS-4 induction by the cytokines and its suppression by were treated with IL-1␤ or TNF-␣ for 1 h and nuclear extracts used siRNAs occurs at the pretranscriptional and signal transduction to determine NF-␬B DNA binding activity by ELISA. IL-1␤-in- levels. Inhibition of MyD88 by overexpression of dominant neg- duced activity of NF-␬B was significantly suppressed in Ras-de- ative protein resulted in down-regulation of other MMPs including ficient chondrocytes (Fig. 4B). Because Ras siRNA along with MMP-13 in the human synovial membrane cultures (38). We have IRAK, MyD88, or TRAF6 has an additive effect on IL-1␤ regu- recently shown that MyD88 knockdown inhibits MMP-13 expres- lation of ADAMTS-4, we investigated whether the deficiency of sion in human chondrocytes, which is the major collagen-cleaving Ras and MyD88, IRAK1, or TRAF6 had stronger impact on the enzyme in arthritis (39). Similar to our results, NF-␬B activation is phosphorylation of IKK␣␤ and I␬B␣. The combined deficiency of blocked in MyD88-knockout cells and by a dominant-negative Ras with MyD88, IRAK, or TRAF6 completely eliminated the splice variant of MyD88 (40, 41). Further, IL-1 induction of phosphorylation of IKK␣␤ and I␬B␣ and increased I␬B␣ stabili- NF-␬B requires IRAK (42). MyD88 is a critical mediator of zation (Fig. 4C), which lead to abrogation of the NF-␬B transcrip- monosodium urate crystal-stimulated gout-like inflammation, as tion factor activity (Fig. 4D). MyD88-deficient mice display reduced inflammation (43, 44). Similarly, MyD88-knockout mice are resistant to serum-, Esche- Discussion richia coli-, or streptococcal cell wall-induced reactive arthritis Aggrecanases or ADAMTSs are the major pharmacological targets (45–47). In agreement with our results with the IRAK-inhibited for blocking structural damage to joint tissues during the course of chondrocytes, IRAK1-deficient mouse embryonic fibroblasts dis- arthritis. In this study, we have shown by genetic knockdown and played a reduced activation of JNK, p38, and NF-␬B by IL-1 (48). antioxidant approaches that ADAMTS-4 induction by the proin- IRAK1-TRAF6 interaction is important for downstream activation flammatory cytokine, IL-1␤ is mediated in part by MyD88, of JNK and NF-␬B (49). Knockout of IRAK-4, an IRAK-1 inter- IRAK1, TRAF6, and NF-␬B as well as by Ras through the pro- acting protein, also impairs IL-1 signaling (50). Thus, inhibition of duction of reactive oxygen species. We further show that any of these critical mediator proteins (MyD88, IRAK, or TRAF6) ADAMTS-4 induction by IL-1␤ could be potently inhibited by the could block the expression of arthritic damage-associated pro- combined knockdown of the individual adaptor proteins and Ras teases, such as ADAMTS-4 and MMP-13. by the respective siRNAs and by antioxidants. Partial inhibition of ADAMTS-4 expression and activity by Lack of inhibition of the respective adaptor protein by negative MyD88, IRAK, or TRAF6 deficiency led us to hypothesize that control siRNA and strong genetic ablation by the respective spe- additional signaling cascades may also be involved in ADAMTS-4 cific siRNAs clearly demonstrated the specificity of RNA interfer- up-regulation by IL-1. Incomplete inhibition of ADAMTS-4 5086 AGGRECANASE-1 REGULATION BY MYD88, IRAK1, TRAF6 AND RAS

gene. The reduced phosphorylation does not appear to be due to total I␬B␣ degradation. Further downstream of adaptor proteins, TRAF6 could recruit TAK1 to phosphorylate IKK and activate NF-␬B. The major collagenolytic enzyme in arthritis, MMP-13, is also a known target of TAK1 (63). Thus, based on our results, we propose the following overlapping signaling pathways for ADAMTS-4 gene induction by IL-1␤. Binding of IL-1␤ to its receptor could recruit the adaptor proteins MyD88, IRAK1, and TRAF6 and through TAK1 could activate IKK␣␤,I␬B␣, and NF- ␬B. IL-1 could also activate Ras and through MyD88-dependent endocytosis of IL-1R, recruitment of Ras and Rac1-dependent NADPH oxidases (Nox2/Nox4 catalytic subunits) to the endosome, NADPH oxidase-mediated production of superoxide, its dis-

mutation into H2O2 spontaneously or by superoxide dismutase. ROS can diffuse out of the endosome and in turn associate with FIGURE 5. Proposed model for the ADAMTS-4 gene induction by IL- TRAF6 and activate IKK and NF-␬B (37, 58). Nox4 overexpres- 1␤. Two parallel cascades involving canonical NF-␬B and alternate Ras-, sion has been found to increase ROS and collagenase in a chon- Rac-, and ROS-dependant pathways are depicted. Individual knockdown of drocytic cell line (29). Alternatively, Ras may be activated inde- MyD88, IRAK1, TRAF6, Ras, or ROS partially inhibits ADAMTS-4. pendent of adaptor proteins in parallel. All these upstream Downloaded from Combined knockdown of components of the two pathways by siRNA or molecules could subsequently activate transcription factor NF-␬B antioxidants more potently inhibits ADAMTS-4 gene expression. SOD, superoxide dismutase; IKK, I␬B kinase; I␬B, inhibitory ␬B; TAK, TGF- and induce its target, the ADAMTS-4 gene (Fig. 5). ␤-activated kinase; NEMO, NF-␬B essential modifier. Disclosures The authors have no financial conflict of interest. mRNA and activity by Ras-specific siRNA and not by nonspecific http://www.jimmunol.org/ control siRNA suggested the implication of Ras as well. More References potent inhibition by Ras and adaptor protein siRNA combinations 1. Roughley, P. J. 2006. The structure and function of cartilage proteoglycans. Eur. further supported synergistic regulation of ADAMTS-4 by these Cell Mater. 12: 92–101. 2. Goldring, M. B., J. R. Birkhead, L. F. Suen, R. Yamin, S. Mizuno, J. Glowacki, mediators. Ras regulation of ADAMTS-4 has never been studied J. L. Arbiser, and J. F. Apperley. 1994. Interleukin-1 ␤-modulated gene expres- before. Because Ras is farnesylated at the C-terminal CAAX box sion in immortalized human chondrocytes. J. Clin. Invest. 94: 2307–2316. 3. Zwerina, J., K. Redlich, K. Polzer, L. Joosten, G. Kronke, J. Distler, A. Hess, in cytosol by farnesyltransferase, the AAX tripeptide cleaved, N. Pundt, T. Pap, O. Hoffmann, et al. 2007. TNF-induced structural joint damage methylated, and then transported to the plasma membrane, farne- is mediated by IL-1. Proc. Natl. Acad. Sci. USA 104: 11742–11747. syltransferase inhibitors have been shown to inhibit MMP-1 and 4. McInnes, I. B., and G. Schett. 2007. Cytokines in the pathogenesis of rheumatoid

arthritis. Nat. Rev. Immunol. 7: 429–442. by guest on September 24, 2021 MMP-13 expression (51, 52). Such inhibitors also reduce severity 5. Tetlow, L. C., D. J. Adlam, and D. E. Woolley. 2001. Matrix metalloproteinase of collagen-induced arthritis (an animal model of RA) by suppress- and proinflammatory cytokine production by chondrocytes of human osteoar- ing Ras and IKK-mediated activation of NF-␬B (53). Previously thritic cartilage: associations with degenerative changes. Arthritis Rheum. 44: 585–594. observed suppression of bone destruction by dominant negative 6. Dayer, J. M., and B. Bresnihan. 2002. Targeting interleukin-1 in the treatment of Ras may be partly due to inhibition of proteases such as rheumatoid arthritis. Arthritis Rheum. 46: 574–578. 7. McCulloch, C. A., G. P. Downey, and H. El Gabalawy. 2006. Signalling plat- ADAMTS-4 and MMPs (54, 55). forms that modulate the inflammatory response: new targets for drug develop- IL-1 can activate Ras (25, 26) and activated Ras through Rac1 ment. Nat. Rev. Drug Discov. 5: 864–876. (56) leads to production of ROS in many cell types including in- 8. Abramson, S. B., and A. Amin. 2002. Blocking the effects of IL-1 in rheumatoid arthritis protects bone and cartilage. Rheumatology 41: 972–980. tracellularly generated H2O2 in the RA synovial fibroblasts and T 9. Fan, Z., S. Soder, S. Oehler, K. Fundel, and T. Aigner. 2007. Activation of lymphocytes (57). The mechanism may involve NADPH oxidase- interleukin-1 signaling cascades in normal and osteoarthritic articular cartilage. mediated production of superoxide and H O (37, 58). IL-1 has Am. J. Pathol. 171: 938–946. 2 2 10. Bondeson, J., S. Wainwright, C. Hughes, and B. Caterson. 2008. The regulation recently been shown to disturb the balance of endogenous oxidant- of the ADAMTS4 and ADAMTS5 aggrecanases in osteoarthritis: a review. Clin. antioxidant enzyme system thus creating oxidative stress in bovine Exp. Rheumatol. 26: 139–145. 11. Porter, S., I. M. Clark, L. Kevorkian, and D. R. Edwards. 2005. The ADAMTS chondrocytes (59). The potent inhibition of ADAMTS-4 mRNA metalloproteinases. Biochem. J. 386: 15–27. expression and activity by the antioxidants Trolox and NDGA sup- 12. Naito, S., T. Shiomi, A. Okada, T. Kimura, M. Chijiiwa, Y. Fujita, T. Yatabe, ports the involvement of ROS production by 5-lipoxygenase or K. Komiya, H. Enomoto, K. Fujikawa, and Y. Okada. 2007. Expression of ADAMTS4 (aggrecanase-1) in human osteoarthritic cartilage. Pathol. Int. 57: NADPH oxidase during IL-1 induction of ADAMTS-4 (31, 58). 703–711. ADAMTS-4 is partly regulated by NF-␬B transcription factor as 13. Struglics, A., S. Larsson, M. A. Pratta, S. Kumar, M. W. Lark, and its induction in OA synovial fibroblasts is inhibited by the over- L. S. Lohmander. 2006. Human osteoarthritis synovial fluid and joint cartilage contain both aggrecanase- and matrix metalloproteinase-generated aggrecan frag- expression of the endogenous NF-␬B inhibitor, I␬B␣ (18, 19, 60). ments. Osteoarthritis Cartilage 14: 101–113. IL-1-stimulated NF-␬B is known to be more active in the super- 14. Glasson, S. S., R. Askew, B. Sheppard, B. Carito, T. Blanchet, H. L. Ma, ␬ C. R. Flannery, D. Peluso, K. Kanki, Z. Yang, et al. 2005. Deletion of active ficial layer of OA cartilage (9). NF- B is considered a redox sig- ADAMTS5 prevents cartilage degradation in a murine model of osteoarthritis. naling-sensitive transcription factor in a cell-type specific manner Nature 434: 644–648. by some but not by others (61). In our experience, several anti- 15. Stanton, H., F. M. Rogerson, C. J. East, S. B. Golub, K. E. Lawlor, C. T. Meeker, C. B. Little, K. Last, P. J. Farmer, I. K. Campbell, et al. 2005. ADAMTS5 is the oxidants inhibit cytokine-induced expression of ADAMTS-4 pos- major aggrecanase in mouse cartilage in vivo and in vitro. Nature 434: 648–652. sibly through Ras-NF-␬B pathway (62). 16. Song, R. H., M. D. Tortorella, A. M. Malfait, J. T. Alston, Z. Yang, E. C. Arner, The observed inhibition of IKK␣␤ and I␬B␣ phosphorylation and D. W. Griggs. 2007. Aggrecan degradation in human articular cartilage explants is mediated by both ADAMTS-4 and ADAMTS-5. Arthritis Rheum. 56: and NF-␬B activity by MyD88, IRAK1, TRAF6, or Ras siRNA 575–585. alone (Fig. 3) or in combination (Fig. 4) suggests that Ras and 17. Little, C. B., C. T. Meeker, S. B. Golub, K. E. Lawlor, P. J. Farmer, S. M. Smith, and A. J. Fosang. 2007. Blocking aggrecanase cleavage in the aggrecan inter- adaptor proteins are early upstream events in the activation of globular domain abrogates cartilage erosion and promotes cartilage repair. NF-␬B and regulation of its downstream target, the ADAMTS-4 J. Clin. Invest. 117: 1627–1636. The Journal of Immunology 5087

18. Chockalingam, P. S., U. Varadarajan, R. Sheldon, E. Fortier, E. R. LaVallie, 43. Chen, C. J., Y. Shi, A. Hearn, K. Fitzgerald, D. Golenbock, G. Reed, S. Akira, E. A. Morris, P. J. Yaworsky, and M. K. Majumdar. 2007. Involvement of protein and K. L. Rock. 2006. MyD88-dependent IL-1 receptor signaling is essential for kinase C␨ in interleukin-1␤ induction of ADAMTS-4 and type 2 nitric oxide gouty inflammation stimulated by monosodium urate crystals. J. Clin. Invest. synthase via NF-␬B signaling in primary human osteoarthritic chondrocytes. Ar- 116: 2262–2271. thritis Rheum. 56: 4074–4083. 44. Liu-Bryan, R., P. Scott, A. Sydlaske, D. M. Rose, and R. Terkeltaub. 2005. Innate 19. LaVallie, E. R., P. S. Chockalingam, L. A. Collins-Racie, B. A. Freeman, immunity conferred by Toll-like receptors 2 and 4 and myeloid differentiation C. C. Keohan, M. Leitges, A. J. Dorner, E. A. Morris, M. K. Majumdar, and factor 88 expression is pivotal to monosodium urate monohydrate crystal-induced M. Arai. 2006. Protein kinase C␨ is up-regulated in osteoarthritic cartilage and is inflammation. Arthritis Rheum. 52: 2936–2946. required for activation of NF-␬B by tumor necrosis factor and interleukin-1 in 45. Choe, J. Y., B. Crain, S. R. Wu, and M. Corr. 2003. Interleukin 1 receptor articular chondrocytes. J. Biol. Chem. 281: 24124–24137. dependence of serum transferred arthritis can be circumvented by toll-like recep- 20. Qureshi, H. Y., R. Ahmad, and M. Zafarullah. 2008. High-efficiency transfection tor 4 signaling. J. Exp. Med. 197: 537–542. of nucleic acids by the modified calcium phosphate precipitation method in chon- 46. Kyo, F., H. Futani, K. Matsui, M. Terada, K. Adachi, K. Nagata, H. Sano, drocytes. Anal. Biochem. 382: 138–140. H. Tateishi, H. Tsutsui, and K. Nakanishi. 2005. Endogenous interleukin-6, but 21. Little, C. B., C. E. Hughes, C. L. Curtis, S. A. Jones, B. Caterson, and not tumor necrosis factor ␣, contributes to the development of toll-like receptor C. R. Flannery. 2002. Cyclosporin A inhibition of aggrecanase-mediated proteo- 4/myeloid differentiation factor 88-mediated acute arthritis in mice. Arthritis glycan catabolism in articular cartilage. Arthritis Rheum. 46: 124–129. Rheum. 52: 2530–2540. 22. Miller, J. A., R. Q. Liu, G. L. Davis, M. A. Pratta, J. M. Trzaskos, and 47. Joosten, L. A., M. I. Koenders, R. L. Smeets, M. Heuvelmans-Jacobs, R. A. Copeland. 2003. A microplate assay specific for the enzyme aggrecanase. M. M. Helsen, K. Takeda, S. Akira, E. Lubberts, F. A. van de Loo, and Anal. Biochem. 314: 260–265. W. B. van den Berg. 2003. Toll-like receptor 2 pathway drives streptococcal cell 23. Pratta, M. A., J. L. Su, M. A. Leesnitzer, A. Struglics, S. Larsson, wall-induced joint inflammation: critical role of myeloid differentiation factor 88. L. S. Lohmander, and S. Kumar. 2006. Development and characterization of a J. Immunol. 171: 6145–6153. highly specific and sensitive sandwich ELISA for detection of aggrecanase-gen- 48. Kanakaraj, P., P. H. Schafer, D. E. Cavender, Y. Wu, K. Ngo, P. F. Grealish, erated aggrecan fragments. Osteoarthritis Cartilage 14: 702–713. S. A. Wadsworth, P. A. Peterson, J. J. Siekierka, C. A. Harris, and 24. Li, H., and X. Lin. 2008. Positive and negative signaling components involved in W. P. Fung-Leung. 1998. Interleukin (IL)-1 receptor-associated kinase (IRAK) TNF␣-induced NF-␬B activation. Cytokine 41: 1–8. requirement for optimal induction of multiple IL-1 signaling pathways and IL-6 25. Palsson, E. M., M. Popoff, M. Thelestam, and L. A. O’Neill. 2000. Divergent production. J. Exp. Med. 187: 2073–2079. Downloaded from roles for Ras and Rap in the activation of p38 mitogen-activated protein kinase 49. Yanagisawa, K., K. Tago, M. Hayakawa, M. Ohki, H. Iwahana, and S. Tominaga. by interleukin-1. J. Biol. Chem. 275: 7818–7825. 2003. A novel splice variant of mouse interleukin-1-receptor-associated kinase-1 26. Caunt, C. J., E. Kiss-Toth, F. Carlotti, R. Chapman, and E. E. Qwarnstrom. 2001. (IRAK-1) activates nuclear factor-␬B (NF-␬B) and c-Jun N-terminal kinase Ras controls tumor necrosis factor receptor-associated factor (TRAF)6-dependent (JNK). Biochem. J. 370: 159–166. induction of nuclear factor-␬b: selective regulation through receptor signaling 50. Suzuki, N., S. Suzuki, G. S. Duncan, D. G. Millar, T. Wada, C. Mirtsos, components. J. Biol. Chem. 276: 6280–6288. H. Takada, A. Wakeham, A. Itie, S. Li, et al. 2002. Severe impairment of inter- 27. Lo, Y. Y., J. A. Conquer, S. Grinstein, and T. F. Cruz. 1998. Interleukin-1 ␤ leukin-1 and Toll-like receptor signaling in mice lacking IRAK-4. Nature 416:

induction of c-fos and collagenase expression in articular chondrocytes: involve- 750–756. http://www.jimmunol.org/ ment of reactive oxygen species. J. Cell Biochem. 69: 19–29. 51. Lee, S. K., J. M. Kim, M. Y. Lee, K. H. Son, Y. I. Yeom, C. H. Kim, Y. Shin, 28. Tiku, M. L., Y. P. Yan, and K. Y. Chen. 1998. Hydroxyl radical formation in J. S. Koh, D. C. Han, and B. M. Kwon. 2006. Conﬁrmation of a linkage between chondrocytes and cartilage as detected by electron paramagnetic resonance spec- H-Ras and MMP-13 expression as well as MMP-9 by chemical genomic ap- troscopy using spin trapping reagents. Free Radical Res. 29: 177–187. proach. Int. J. Cancer 118: 2172–2181. 29. Grange, L., M. V. Nguyen, B. Lardy, M. Derouazi, Y. Campion, C. Trocme, 52. Abeles, A. M., N. Marjanovic, J. Park, M. Attur, E. S. Chan, H. E. Al Mussawir, M. H. Paclet, P. Gaudin, and F. Morel. 2006. NAD(P)H oxidase activity of Nox4 M. Dave, M. C. Fisher, S. A. Stuchin, S. B. Abramson, and M. H. Pillinger. 2007. in chondrocytes is both inducible and involved in collagenase expression. Anti- Protein isoprenylation regulates secretion of matrix metalloproteinase 1 from oxid. Redox. Signal 8: 1485–1496. rheumatoid synovial ﬁbroblasts: effects of statins and farnesyl and geranylgeranyl 30. Heo, J., and S. L. Campbell. 2006. Ras regulation by reactive oxygen and nitro- transferase inhibitors. Arthritis Rheum. 56: 2840–2853. gen species. Biochemistry 45: 2200–2210. 53. Na, H. J., S. J. Lee, Y. C. Kang, Y. L. Cho, W. D. Nam, P. K. Kim, K. S. Ha, 31. Bonizzi, G., J. Piette, S. Schoonbroodt, R. Greimers, L. Havard, M. P. Merville, H. T. Chung, H. Lee, Y. G. Kwon, J. S. Koh, and Y. M. Kim. 2004. Inhibition

and V. Bours. 1999. Reactive oxygen intermediate-dependent NF-␬B activation of farnesyltransferase prevents collagen-induced arthritis by down-regulation of by guest on September 24, 2021 by interleukin-1␤ requires 5-lipoxygenase or NADPH oxidase activity. Mol. Cell inflammatory gene expression through suppression of p21(ras)-dependent NF-␬B Biol. 19: 1950–1960. activation. J. Immunol. 173: 1276–1283. 32. Brzezinska-Slebodzinska, E., and B. Pietras. 1997. The protective role of some 54. Yamamoto, A., A. Fukuda, H. Seto, T. Miyazaki, Y. Kadono, Y. Sawada, antioxidants and scavengers on the free radicals-induced inhibition of the liver I. Nakamura, H. Katagiri, T. Asano, Y. Tanaka, et al. 2003. Suppression of iodothyronine 5Ј-monodeiodinase activity and thiols content. J. Physiol. Phar- arthritic bone destruction by adenovirus-mediated dominant-negative Ras gene macol. 48: 451–459. transfer to synoviocytes and osteoclasts. Arthritis Rheum. 48: 2682–2692. 33. Mazor, D., L. Greenberg, D. Shamir, D. Meyerstein, and N. Meyerstein. 2006. 55. Sun, Y., L. Wenger, C. E. Brinckerhoff, R. R. Misra, and H. S. Cheung. 2002. Antioxidant properties of bucillamine: possible mode of action. Biochem. Bio- Basic calcium phosphate crystals induce matrix metalloproteinase-1 through the phys. Res. Commun. 349: 1171–1175. Ras/mitogen-activated protein kinase/c-Fos/AP-1/metalloproteinase 1 pathway: 34. Del Carlo, M., D. Schwartz, E. A. Erickson, and R. F. Loeser. 2007. Endogenous involvement of transcription factor binding sites AP-1 and PEA-3. J. Biol. Chem. production of reactive oxygen species is required for stimulation of human ar- 277: 1544–1552. ticular chondrocyte matrix metalloproteinase production by fibronectin frag- 56. Finkel, T. 2006. Intracellular redox regulation by the family of small GTPases. ments. Free Radical Biol. Med. 42: 1350–1358. Antioxid. Redox. Signal 8: 1857–1863. 35. Ramasamy, S., G. R. Drummond, J. Ahn, M. Storek, J. Pohl, S. Parthasarathy, 57. Remans, P. H., M. van Oosterhout, T. J. Smeets, M. Sanders, W. M. Frederiks, and D. G. Harrison. 1999. Modulation of expression of endothelial nitric oxide K. A. Reedquist, P. P. Tak, F. C. Breedveld, and J. M. van Laar. 2005. Intracel- synthase by nordihydroguaiaretic acid, a phenolic antioxidant in cultured endo- lular free radical production in synovial T lymphocytes from patients with rheu- thelial cells. Mol. Pharmacol. 56: 116–123. matoid arthritis. Arthritis Rheum. 52: 2003–2009. 36. Roman-Blas, J. A., and S. A. Jimenez. 2006. NF-␬B as a potential therapeutic 58. Li, Q., M. M. Harraz, W. Zhou, L. N. Zhang, W. Ding, Y. Zhang, T. Eggleston, target in osteoarthritis and rheumatoid arthritis. Osteoarthritis Cartilage 14: C. Yeaman, B. Banfi, and J. F. Engelhardt. 2006. Nox2 and Rac1 regulate H2O2- 839–848. dependent recruitment of TRAF6 to endosomal interleukin-1 receptor complexes. 37. Gloire, G., S. Legrand-Poels, and J. Piette. 2006. NF-␬B activation by reactive Mol. Cell Biol. 26: 140–154. oxygen species: fifteen years later. Biochem. Pharmacol. 72: 1493–1505. 59. Mathy-Hartert, M., L. Hogge, C. Sanchez, G. Deby-Dupont, J. M. Crielaard, and 38. Sacre, S. M., E. Andreakos, S. Kiriakidis, P. Amjadi, A. Lundberg, G. Giddins, Y. Henrotin. 2008. Interleukin-1␤ and interleukin-6 disturb the antioxidant en- M. Feldmann, F. Brennan, and B. M. Foxwell. 2007. The Toll-like receptor zyme system in bovine chondrocytes: a possible explanation for oxidative stress adaptor proteins MyD88 and Mal/TIRAP contribute to the inflammatory and generation. Osteoarthritis Cartilage 16: 756–763. destructive processes in a human model of rheumatoid arthritis. Am. J. Pathol. 60. Bondeson, J., S. Lauder, S. Wainwright, N. Amos, A. Evans, C. Hughes, 170: 518–525. M. Feldmann, and B. Caterson. 2007. Adenoviral gene transfer of the endogenous 39. Ahmad, R., J. Sylvester, and M. Zafarullah. 2007. MyD88, IRAK1 and TRAF6 inhibitor I␬B␣ into human osteoarthritis synovial fibroblasts demonstrates that knockdown in human chondrocytes inhibits interleukin-1-induced matrix metal- several matrix metalloproteinases and aggrecanases are nuclear factor-␬B-depen- loproteinase-13 gene expression and promoter activity by impairing MAP kinase dent. J. Rheumatol. 34: 523–533. activation. Cell Signal. 19: 2549–2557. 61. Hayakawa, M., H. Miyashita, I. Sakamoto, M. Kitagawa, H. Tanaka, H. Yasuda, 40. Adachi, O., T. Kawai, K. Takeda, M. Matsumoto, H. Tsutsui, M. Sakagami, M. Karin, and K. Kikugawa. 2003. Evidence that reactive oxygen species do not K. Nakanishi, and S. Akira. 1998. Targeted disruption of the MyD88 gene results mediate NF-␬B activation. EMBO J. 22: 3356–3366. in loss of IL-1- and IL-18-mediated function. Immunity 9: 143–150. 62. El Mabrouk, M., R. Ahmad, J. Sylvester, and M. Zafarullah. 2006. Curcumin, 41. Janssens, S., K. Burns, E. Vercammen, J. Tschopp, and R. Beyaert. 2003. nordihydroguiaretic acid, quercetin and resveratrol inhibit interleukin-1-induced MyD88S, a splice variant of MyD88, differentially modulates NF-␬B- and AP- ADAMTS-4 (aggrecanase-1) gene expression in articular chondrocytes: natural 1-dependent gene expression. FEBS Lett. 548: 103–107. products as potential anti-arthritic agents. Pharmacologyonline 3: 601–610. 42. Croston, G. E., Z. Cao, and D. V. Goeddel. 1995. NF-␬B activation by interleu- 63. Klatt, A. R., G. Klinger, O. Neumuller, B. Eidenmuller, I. Wagner, T. Achenbach, kin-1 (IL-1) requires an IL-1 receptor-associated protein kinase activity. J. Biol. T. Aigner, and E. Bartnik. 2006. TAK1 downregulation reduces IL-1␤ induced Chem. 270: 16514–16517. expression of MMP13, MMP1 and TNF-␣. Biomed. Pharmacother. 60: 55–61. The Journal of Immunology

Corrections

Ahmad, R., J. Sylvester, M. Ahmad, and M. Zafarullah. 2009. Adaptor proteins and Ras synergistically regulate IL-1-induced ADAMTS-4 expression in human chondrocytes. J. Immunol. 182: 5081–5087.

ADAMTS was incorrectly deﬁned as “a dystrophin and metalloproteinase with thrombospondin motif” in the following locations: the abbreviations footnote, the ﬁrst sentence of the Abstract, and the second sentence of the second paragraph of the Introduction.

The correct term is, “a disintegrin and metalloproteinase with thrombospondin motif.”

The authors thank Dr. Amanda Fosang, University of Melbourne (Victoria, Australia), for pointing out this error. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1090045